Introduction

Bacterial secondary metabolites often exhibit remarkable bioactivities and have significantly contributed to human society through their application in pharmaceuticals, agrochemicals and food additives. In particular, since the discovery of actinomycins and streptomycin [1, 2], numerous bioactive secondary metabolites have been identified from actinomycetes [3–5]. However, the continuous exploration of actinomycetes has made the discovery of new natural products increasingly challenging.

Meanwhile, genomic studies have revealed that actinomycetes harbor numerous secondary metabolite biosynthetic gene clusters (SM-BGCs) [6–8], far exceeding the total number of actinobacterial secondary metabolites identified to date. Thus, the majority of SM-BGCs remain cryptic under standard culture conditions, and activating these untapped actinobacterial SM-BGCs is expected to facilitate the discovery of new bioactive compounds. In fact, new secondary metabolites have been successfully identified through the activation of cryptic SM-BGCs using heterologous expression [9, 10], the application of chemical elicitors [11, 12], and genetic manipulation [13].

Furthermore, secondary metabolism involving elements that are rarely found in natural products is also noteworthy. For example, some actinomycetes possess specific SM-BGCs and produce organofluorine [14–16] and organoselenium metabolites [17], representing an extremely rare class of natural products. However, a large portion of these SM-BGCs is likely overlooked, even if they are functional, because conventional culture media typically lack the necessary elements for their biosynthesis.

Under these circumstances, we have continued our efforts to discover new natural products by utilizing untapped actinobacterial SM-BGCs. In this review, we first describe our studies on the activation of actinobacterial SM-BGCs through co-culture with mycolic acid-containing bacteria (MACB). In the latter part, we present our recent findings on actinobacterial secondary metabolism related to organoarsenical natural products.

Combined-culture: effective activation of actinobacterial SM-BGCs through co-culture of actinomycetes with MACB

Co-culture is also an effective approach for activating cryptic SM-BGCs, and several co-culture studies have successfully led to the isolation of new microbial natural products [18–20]. It is a simple and broadly applicable method that does not require genetic manipulation or expensive reagents. However, conventional co-culture approaches often require extensive screening to identify suitable microbial combinations. To address this limitation, Onaka et al. proposed an efficient co-culture strategy using MACB as a partner to activate cryptic SM-BGCs in actinomycetes, which they termed "combined-culture" [21].

Initially, they conducted a screening of bacteria capable of activating actinobacterial SM-BGCs using Streptomyces lividans TK23 as an indicator, which produces pigmented antibiotics (actinorhodin and undecylprodigiosin) in response to specific stimulation. As a result, Tsukamurella pulmonis TP-B0596 (T. pulmonis) was identified as an effective inducer of pigmented antibiotics production in S. lividans TK23, both on agar plates (Fig. 1a) and in liquid cultures (Fig. 1b) [21].

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

a Production of pigmented antibiotics in Streptomyces lividans TK23 (S. lividans) was induced by co-culturing with Tsukamurella pulmonis TP-B0596 (T. pulmonis) on an agar plate. b Production of pigmented antibiotics in S. lividans was also induced in liquid culture through co-cultivation with T. pulmonis. c Chemical structures of alchivemycins (1 and 2), with the 2H-tetrahydro-4,6-dioxo-1,2-oxazine moiety highlighted in red

The genus Tsukamurella belongs to MACB, a group of bacteria characterized by the presence of mycolic acid on their cell surface. Notably, other MACB genera, such as Corynebacterium, Dietzia, and Rhodococcus, also induced pigment production in S. lividans TK23. The importance of mycolic acid was further supported by the finding that Corynebacterium glutamicum Δpks13, a mutant lacking the mycolic acid biosynthetic gene, lost its inducing activity [21].

More importantly, MACB not only induced pigment production in S. lividans TK23 but also broadly activated SM-BGCs in various actinomycetes. Onaka et al. reported that co-culturing T. pulmonis with 112 strains of soil-derived Streptomyces resulted in metabolic profile changes in 97 strains, among which 35 strains exhibited enhanced secondary metabolite production [21]. In particular, two novel polyketides featuring a unique 2H-tetrahydro-4,6-dioxo-1,2-oxazine moiety, named alchivemycins A (1) and B (2), were isolated from Streptomyces sp. TP-A0867 when co-cultured with T. pulmonis (Fig. 1c) [21–23]. Alchivemycins exhibited potent and selective antibacterial activity against Kocuria rhizophila ATCC 9341 (formerly Micrococcus luteus ATCC 9341), with MIC values of 0.03 µg/ml for 1 and 0.004 µg/ml for 2 [23].

Application of combined-culture strategy to Streptomyces strains

Encouraged by the discovery of alchivemycins, which possess unique chemical structures and potent bioactivity, we aimed to obtain new secondary metabolites through the combined-culture strategy. Initially, we targeted Streptomyces strains, a representative genus of actinomycetes that has produced the majority of actinobacterial secondary metabolites [3, 24]. In our combined-culture screening, T. pulmonis was consistently used as the MACB, and changes in the metabolic profile were monitored by HPLC equipped with a diode array detector (HPLC–DAD). Thus far, we have isolated eleven secondary metabolites, including seven new compounds, from four Streptomyces strains as described below (3−13, Fig. 2a).

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

a Chemical structures of secondary metabolites (3–13) obtained from Streptomyces strains through our combined-culture screening. b Chemical synthesis of possible stereoisomers of 6 and 7. The figure illustrates the case in which (R)-14 was used as the starting material. c Chemical structure of tripartilactam (17), as originally proposed by Oh et al. in 2012 [33]

Streptomyces cinnamoneus NBRC 13823 was found to produce three pigmented metabolites, whose production was significantly enhanced in the presence of MACB [25]. Based on chromatographic purification and spectroscopic analysis, one of these metabolites was identified as a new indolocarbazole alkaloid, named arcyriaflavin E (3), while the remaining two were the known indolocarbazole alkaloids arcyriaflavin A (4) and BE-13793C (5) [26, 27]. 3 and 5 exhibited moderate cytotoxicity against P388 murine leukemia cells, with IC50 values of 39 and 33 µM, respectively, while 4 showed no cytotoxicity up to 100 µM [25].

Streptomyces sp. CJ-5, which we isolated from a soil sample collected in Isumi City (Chiba, Japan), produced three unknown metabolites exclusively in the presence of MACB. Our detailed spectroscopic analysis revealed that all of these metabolites are new butanolides, named chojalactones A–C (6−8) [28]. The relative configurations of 6 and 7 were determined by NMR analysis, while their absolute configurations were established through the chemical synthesis of all possible enantiomers from 3-methyl-γ-butyrolactone (14) (Fig. 2b).

Initially, 14 was condensed with triene acid chloride (15) via Ti-crossed Claisen condensation using N-methylimidazole as a catalyst [29], yielding the β-ketolactone (16). 16 was then subjected to Pd/C-catalyzed direct α-oxidization under an O2 atmosphere [30] to obtain the desired stereoisomers: (2R, 3S) and (2S, 3S) enantiomers from (3R)-16, and (2S, 3R) and (2R, 3R) enantiomers from (3S)-16. By comparing the retention times of the natural products with those of synthetic standards using chiral-phase HPLC, the absolute configurations of 6 and 7 were determined to be (2R, 3S) and (2R, 3R), respectively. The absolute configuration at the C-2 position of 8 was deduced to be R based on its consistency with 6 and 7. Finally, 6 and 7 exhibited moderate cytotoxicity against P388 murine leukemia cells, with IC50 values of 28 and 18 µM, respectively.

It was found that Streptomyces sp. NZ-6, which we isolated from a soil sample collected in Niiza City (Saitama, Japan), produces three metabolites in a MACB-dependent manner. Two of them were identified as new polyene macrolactams with a common [5, 23]-bicyclic skeleton and were named niizalactams A (9) and B (10) [31]. The relative and absolute configurations of 9 and 10 were fully determined based on exhaustive NMR analysis of the natural products and their synthetic derivatives, including the application of the modified Mosher’s method to secondary alcohols [32].

The NMR spectrum of the remaining metabolite closely matched that of tripartilactam (17), a polyene macrolactam with an [4, 8, 18]-tricyclic skeleton, which was isolated from Streptomyces sp. SNA112 by Oh et al. in 2012 (Fig. 2c) [33]. However, based on our thorough NMR analysis, the presence of an [6, 6, 18]-tricyclic skeleton was supported, leading to its designation as niizalactam C (11). Recently, Oh et al. conducted a re-investigation of the chemical structure of 17, including a 13C–13C COSY analysis of a 13C-labeled derivative of 17 [34]. In their study, the original structure of 17, proposed in 2012, was revised to be identical to 11, as proposed by our group.

Along with the above examples, we also found that Streptomyces sp. TAKO-2, which we isolated from a soil sample collected in Tako Town (Chiba, Japan), produced compounds with absorption in the long-wavelength region (> 400 nm) when co-cultured with MACB [35]. The spectral data of two of these metabolites matched those of the known aromatic polyketides, julichrome Q8,8 (12) [36] and julichrome Q6 (13) [37]. Considering the similarity of their UV–vis spectra, the remaining metabolites were also inferred to be related analogs to 12 and 13.

Application of combined-culture screening to non-Streptomyces actinomycetes

As describe above, our combined-culture screening initially targeted Streptomyces strains; however, we later shifted our focus to other actinomycete genera. Non-Streptomyces actinomycetes, often referred to as rare actinomycetes, have been relatively unexplored for secondary metabolites compared to Streptomyces. However, recent genomic studies of non-Streptomyces genera have revealed the presence of SM-BGCs comparable to or even exceeding those in Streptomyces [7, 8]. In fact, a considerable number of new bioactive secondary metabolites have been identified from these genera [5, 38]. Thus far, through combined-culture screening, we have identified nine new secondary metabolites from four non-Streptomyces actinomycetes (18–26, Fig. 3a).

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

a Chemical structures of secondary metabolites (18–26) obtained from non-Streptomyces strains through our combined-culture screening. b 18 spontaneously converts to 19 in DMSO-d6 solution, accompanied by double bond isomerization. c. Non-enzymatic isomerization of 21, leading to the formation of the 3-methyl-isoxazolidinone scaffold in 20

The genus Umezawaea is a relatively new actinomycete genus proposed in 2007 [39], and no reports on its secondary metabolites had been published until we started the combined-culture screening. Umezawaea sp. RD066910 does not produce any notable metabolites under pure-culture conditions; however, we found that it produced two metabolites with similar UV spectra when co-cultured with MACB. Spectral analysis of the purified metabolites revealed that both are new polycyclic tetramate macrolactams and represent the first secondary metabolites identified from the genus Umezawaea. We named them umezawamides A (18) and B (19), which are structural isomers differing only in the position of the carbon–carbon double bonds [40].

During NMR analysis, it was found that 18 gradually converted to 19 in DMSO-d6 at room temperature, suggesting that 19 is an artifact generated from 18 (Fig. 3b). The absolute configurations of 18 and 19, except at the C-16 position, were fully elucidated by combining NMR analysis, conformational search using the Merck Molecular Force Field (MMFF), and ECD calculations. 18 and 19 exhibited cytotoxicity against P388 murine leukemia cells, with IC50 values of 3.7 and 4.8 µM, respectively. Additionally, 18 showed antifungal activity against Candida albicans, whereas 19 did not [40].

Catenuloplanes sp. RD067331 exhibited a significant increase in the production of two metabolites when cultured in the presence of MACB. Spectral analysis revealed that these metabolites are new heterocyclic peptides, which we named catenulobactins A (20) and B (21) [41]. Total acid hydrolysis, followed by chiral-phase GC–MS analysis of 21, revealed that N5-hydroxyornithine is in the d-form, and that both 5-methyl-oxazolines (MeOzn) are derived from l-threonine, leading to the complete determination of its absolute configuration.

20 is a structural isomer of 21, featuring a unique 3-methyl-isoxazolidinone ring in place of one of the MeOzn moieties. Considering the biosynthetic mechanism of acinetobactin proposed by Wencewicz et al. [42], the 3-methyl-isoxazolidinone ring in 20 is presumed to be generated through an intramolecular rearrangement of 21, involving the MeOzn moiety and the adjacent N–OH group (Fig. 3c). Notably, we found that 21 spontaneously converted to 20 when incubated in phosphate buffer (pH 7.0) at room temperature [41]. Finally, we found that 21 exhibits Fe(III)-chelating activity as well as cytotoxicity against P388 murine leukemia cells, with an IC50 value of 22.4 µM, whereas 20 exhibited neither activity [41].

Micromonospora wenchangensis HEK-797, isolated from lake sediment at Hegura Island (Ishikawa, Japan), produced two metabolites specifically in the presence of MACB. NMR analysis of the purified metabolites revealed that one possesses a unique [5, 5, 6, 16]-tetracyclic skeleton, which we named dracolactam A (22), while the other is a new bicyclic lactam featuring a THF ring, named dracolactam B (23) [43]. The relative and absolute configurations of 22 and 23 were determined through NOESY analysis and a careful evaluation of vicinal coupling constants of the natural products, combined with detailed NMR analysis of the corresponding acetonide derivatives and the application of the modified Mosher’s method.

Epoxidation-mediated intramolecular cyclization of polyene macrolactams triggered by combined-culture

Considering the positions of functional groups, 22 and 23 were presumed to originate from a common 26-membered polyene macrolactam (27), whose planar structure is identical to those of micromonolactam [44] and isomicromonolactam [45] (Fig. 4a). In our proposed mechanism, the double bond between C-22 and C-23 of 27 is initially epoxidized, generating an unstable intermediate, which subsequently undergoes intramolecular cyclization.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

a Proposed biosynthetic pathways of dracolactams (22 and 23) and mirilactams C–E (24–26), involving epoxidation-mediated intramolecular cyclization of 26-membered macrolactam precursors. b Proposed biosynthetic pathway of niizalactam A (9) from a 26-membered sceliphrolactam-type precursor (28). Niizalactam B (10) is also presumed to be biosynthesized from the C-10 deoxy derivative of 28. c Chemical structures of ciromicins A (29) and B (30). d Chemical structures of heronamides (32–35)

If the amide nitrogen directly attacks the epoxide ring at C-22, the resulting bicyclic compound undergoes further intramolecular Diels–Alder reaction, forming 22 (Fig. 4a, Path A). Meanwhile, if the hydroxyl group at C-11 attacks the C-14 carbon, inducing a 1,10-addition accompanied by epoxide ring opening, 23 is formed (Fig. 4a, Path B). Notably, M. wenchangensis HEK-797 produced a metabolite with a UV–Vis spectrum and molecular formula identical to 27 under both pure-culture and combined-culture conditions, suggesting that MACB specifically activates the epoxidation of 27 rather than its biosynthesis [43].

A similar biosynthetic strategy can also be applied to 9 and 10, in which a 26-membered polyene macrolactam (28), sharing the same planar structure as sceliphrolactam [46], or its C-10 deoxy derivative, is presumed to be the precursor (Fig. 4b). It was assumed that epoxidation at the C-22 and C-23 double bond, followed by nucleophilic pyrrolidinol formation, leads to the generation of 9 and 10 (Fig. 4b). Additionally, Bachmann et al. reported that the rare actinomycete Nocardiopsis sp. FU40 ΔApoS produced ciromicins A (29) and B (30) (Fig. 4c) strictly in the presence of MACB (Rhodococcus wratislaviensis) [47]. 29 and 30, which contain the pyrrolidinol moiety, are proposed to be biosynthesized through the epoxidation of a 22-membered glycosylated polyene macrolactam, followed by intramolecular cyclization [47].

Based on these findings, we hypothesized that applying combined-culture could activate cryptic epoxidation-mediated cyclization pathways in polyene macrolactam-producing actinomycetes. To validate this hypothesis, we focused on Actinosynnema mirum NBRC 14064 (A. mirum), which is known to produce a 26-membered polyene macrolactam named mirilactam A (31) (Fig. 4a) [48].

We confirmed that A. mirum produces 31 under monoculture conditions, and as expected, specifically produced three additional metabolites when co-cultured with MACB [49]. Detailed structural analysis of the purified metabolites revealed that they are all new compounds derived from 31, which we named mirilactams C–E (24−26, Fig. 3a) [49]. The biosynthetic mechanism of 24 and 25 is similar to that of 22 and 23. Meanwhile, in the biosynthesis of 26, it is likely that the hydroxyl group at C-10 in the epoxidized intermediate attacks the C-14 carbon, leading to the formation of a tetrahydropyran ring (Fig. 4a, Path C).

Polycyclic metabolites, likely generated by the epoxidation of polyene macrolactams, have also been found in the mono-culture of actinomycetes. Heronamide A (32) is a [5, 5, 6, 10]-tetracyclic metabolite containing a pyrrolidinol moiety, identified from the marine-derived Streptomyces sp. CMB-M0406 (Fig. 4d) [50, 51], while heronamide D (33), which shares the same tetracyclic skeleton as 32, was identified from the deep-sea-derived Streptomyces sp. SCSIO 03032 (Fig. 4d) [52].

Strains CMB-M0406 and SCSIO 03032 also produce 20-membered polyene macrolactams, named heronamides C (34) and F (35), respectively [50, 52, 53] (Fig. 4d), which are likely precursors of 32 and 33. Notably, the production of 32 significantly increased when artificial seawater was added to the culture medium, whereas no difference was observed in the production of 34 between saline and non-saline conditions [50]. Additionally, the production of 33 was observed in the medium containing artificial seawater [52], suggesting that salt stress induce the epoxidation of polyene macrolactams in certain actinomycetes.

To the best of my knowledge, no epoxidase genes have been clearly identified within the polyene macrolactam BGCs. Thus, it is possible that an epoxidase gene located outside the polyene macrolactam BGC is activated during co-cultivation with MACB. On the other hand, it has been reported that 34 undergoes a non-enzymatic conversion to 32 in DMSO at room temperature [53], raising the possibility that the MACB-induced epoxidation could occur through a non-enzymatic mechanism, such as the generation of reactive oxygen species.

Identification and biosynthetic studies of bisenarsan: a novel organoarsenic secondary metabolite produced by model actinomycetes

Organoarsenic compounds containing C–As bonds are quite rare among bacterial natural products. A representative example is a series of simple metabolites biosynthesized through the sequential methylation of arsenite [As(III)], often followed by oxidation, a process widely recognized as a detoxification mechanism for highly toxic inorganic arsenic [54–56]. In other examples, several cyanobacteria have been reported to produce oxo-arsenosugars [57, 58], while a potent glutamine synthetase inhibitor, arsinothricin, has been identified from the terrestrial bacterium Burkholderia gladioli GSRB05 [59, 60].

In the biosynthesis of these organoarsenic natural products in bacteria, all C–As bonds are formed by S-adenosylmethionine (SAM)-dependent enzymes, among which SAM-dependent arsenic methyltransferases are widely distributed in bacterial genomes [54–56]. In addition, recent studies have shown that radical SAM enzymes are also involved in C–As bond formation. The adenosylation of arsenic in oxo-arsenosugar biosynthesis [61] and the 3-amino-3-carboxypropyl radical transfer to the arsenic atom in arsinothricin biosynthesis [62] are both catalyzed by radical SAM enzymes.

Despite their high capacity for secondary metabolite production, actinomycetes have long been known to produce only simple methylated arsenicals [56]. However, in 2016, Cruz-Morales et al. reported that the two model actinomycetes Streptomyces lividans 1326 and Streptomyces coelicolor A3(2) produce an unknown organoarsenic metabolite (36, C14H27AsO5) [63]. They further proposed that a conserved SM-BGC among these strains, which we later termed the bsn cluster, is responsible for the biosynthesis of 36 (Fig. 5a) [63].

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

a Schematic representation of the SM-BGC for 36 in Streptomyces lividans 1326 (bsn cluster). b Chemical structure of 36, along with its proposed components (37 and 38). c Proposed biosynthetic pathway of 36 from As(V)

The bsn cluster includes genes involved in arsenic resistance and arsenic-dependent transcriptional regulation. In fact, it has been reported that the transcription of several genes within the bsn cluster is significantly upregulated under exposure to arsenate [As(V)] [63]. More importantly, the bsn cluster does not contain any genes encoding SAM-dependent enzymes. However, it has been reported that disruption of the 5-enolpyruvylshikimate-3-phosphate (EPSP) synthase homolog (bsnM) completely abolished the production of 36 [63].

Based on these findings, 36 was expected to be an organoarsenic natural product biosynthesized through C–As bond formation without the involvement of SAM-dependent enzymes. Therefore, we aimed to determine the chemical structure of 36. Initially, S. lividans 1326 was mass-cultured in a medium containing As(V) as the sole arsenic source, but only a trace amount of 36 (< 0.1 mg) was obtained [64]. However, 1H NMR and HR-MS/MS analysis of the trace purified material suggested that 36 is an O-acyl derivative of (2-hydroxyethyl)arsonic acid (2-HEA, 37) (Fig. 5b).

In addition, Wang et al. reported that heterologous expression of the polyketide synthase (PKS) genes (bsnPKS and bsnFB, Fig. 5a) resulted in the production of 2,4,6-trimethyl-2-nonenoic acid (2,4,6-TMNA, 38) [65], suggesting that the acyl group of 36 is derived from 38 (Fig. 5b). Notably, when chemically synthesized 37 was added to the culture medium, the production of 36 in S. lividans 1326 significantly increased. Finally, NMR re-analysis of the purified metabolite confirmed that the chemical structure of 36 matched our initial proposal, and we named it bisenarsan (Fig. 5b) [64].

Through gene disruption in S. lividans 1326, we revealed that, in addition to bsnM (EPSP synthase) [63], genes encoding phosphoglycerate mutase (bsnN) and phosphonopyruvate decarboxylase (bsnJ) are also required for the biosynthesis of 36 (Fig. 5a) [64]. Furthermore, supplementation with chemically synthesized 37 in the culture medium of all the disruptants of bsnM, bsnN, and bsnJ restored the production of 36, suggesting that these genes are responsible for the biosynthesis of 37 [64].

Based on these observation as well as the predicted functions of bsn genes, we propose the biosynthetic pathway of 36 as follows (Fig. 5c). Initially, the pyruvate group in EPSP, presumably derived from the shikimate pathway, is transferred to As(V) by BsnM, resulting in the formation of arsenoenolpyruvate (AEP, 39). 39 is then converted to arsonopyruvate (AnPy, 40) by BsnN, accompanied by C–As bond formation, and subsequently decarboxylated to arsonoacetoaldehyde (AnAA, 41) by BsnJ. 41 is presumed to be reduced to 37 by reductases encoded within the bsn cluster or by an orphan reductase in S. lividans 1326. Finally, 38, bound to the acyl carrier protein of the PKS module (BsnPKS and BsnFB), is thought to undergo nucleophilic attack by 37, leading to the release of 36. It has been reported that the expression of bsnPKS is upregulated upon exposure to As(V) [63], suggesting that ester bond formation between 37 and 38 is part of the genuine biosynthetic pathway rather than a shunt pathway.

The biological function of 36 remains unclear; however, we have previously shown that its toxicity is significantly lower than that of As(V) [64], suggesting that the biosynthesis of 36 may serve as a detoxification mechanism for inorganic arsenic. Additionally, we have demonstrated that 37 exhibits stronger antibacterial activity against Staphylococcus aureus than As(V) [64], raising the possibility that 36 is produced as “Trojan horse” antibiotic, with 37 serving as the active form.

Intriguingly, the gene set bsnMN, responsible for the conversion of inorganic arsenic [As(V)] to 40, is widely distributed among actinobacterial genomes, whereas the surrounding regions exhibit considerable genetic diversity [63, 64].

Conclusion

As described earlier, we identified a total of 20 actinobacterial secondary metabolites, including 16 new compounds, from Streptomyces strains (3−13, Fig. 2a) and non-Streptomyces genera (18−26, Fig. 3a) by activating cryptic SM-BGCs through co-culture with MACB (combined-culture). The continued application of the combined-culture strategy to other actinomycetes is expected to facilitate the discovery of additional new bioactive natural products. In fact, the number of new secondary metabolites obtained through this strategy has continued to increase in recent years [66–69].

Furthermore, we isolated and determined the chemical structure of a novel organoarsenic metabolite, named bisenarsan (36), produced by two model actinomycetes (Fig. 5b). Additionally, our study provided significant insight into the biosynthesis of 36, suggesting that C–As bond in 36 is likely generated through a novel pathway that does not involve SAM-dependent enzymes (Fig. 5c). Notably, the bsnMN homologs, presumably responsible for the generation of AnPy (40) from As(V), are widely distributed among actinobacterial genomes, and genome mining targeting these homologs is expected to facilitate the discovery of new organoarsenic metabolites that share 40 as a common precursor.

Overall, a large number of untapped secondary metabolic pathways still exist in actinomycetes, and we continue to explore and utilize these pathways to discover new bioactive secondary metabolites.